Using Natural Language Processing to Extract and Classify Symptoms Among Patients with Thyroid Dysfunction

被引:1
作者
Hwang, Sy [1 ]
Reddy, Sujatha [1 ]
Wainwright, Katherine [1 ]
Schriver, Emily [1 ]
Cappola, Anne [1 ]
Mowery, Danielle [1 ]
机构
[1] Univ Penn, Philadelphia, PA 19104 USA
来源
MEDINFO 2023 - THE FUTURE IS ACCESSIBLE | 2024年 / 310卷
基金
美国国家卫生研究院;
关键词
Natural language processing; machine learning; electronic health records; HYPERTHYROIDISM; OLDER; SIGNS;
D O I
10.3233/SHTI231038
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
In the United States, more than 12% of the population will experience thyroid dysfunction. Patient symptoms often reported with thyroid dysfunction include fatigue and weight change. However, little is understood about the relationship between these symptoms documented in the outpatient setting and ordering patterns for thyroid testing among various patient groups by age and sex. We developed a natural language processing and deep learning pipeline to identify patient-reported outcomes of weight change and fatigue among patients with a thyroid stimulating hormone test. We built upon prior works by comparing 5 open-source, Bidirectional Encoder Representations from Transformers (BERT) to determine which models could accurately identify these symptoms from clinical texts. For both fatigue (f) and weight change (wc), Bio_ClinicalBERT achieved the highest F1-score (f: 0.900; wc: 0.906) compared BERT (f: 0.899; wc: 0.890), DistilBERT (f: 0.852; wc: 0.912), Biomedical RoBERTa (f: 0.864; wc: 0.904), and PubMedBERT (f: 0.882; wc: 0.892).
引用
收藏
页码:614 / 618
页数:5
相关论文
共 8 条
[1]  
American Thyroid Association, 2022, General Information/Press Room
[2]   Older Subjects with Hyperthyroidism Present with a Paucity of Symptoms and Signs: A Large Cross-Sectional Study [J].
Boelaert, K. ;
Torlinska, B. ;
Holder, R. L. ;
Franklyn, J. A. .
JOURNAL OF CLINICAL ENDOCRINOLOGY & METABOLISM, 2010, 95 (06) :2715-2726
[3]   Hypothyroid symptoms and the likelihood of overt thyroid failure: a population-based case-control study [J].
Carle, Allan ;
Pedersen, Inge Bulow ;
Knudsen, Nils ;
Perrild, Hans ;
Ovesen, Lars ;
Laurberg, Peter .
EUROPEAN JOURNAL OF ENDOCRINOLOGY, 2014, 171 (05) :593-602
[4]   Evaluation of a deidentification (De-Id) software engine to share pathology reports and clinical documents for research [J].
Gupta, D ;
Saul, M ;
Gilbertson, J .
AMERICAN JOURNAL OF CLINICAL PATHOLOGY, 2004, 121 (02) :176-186
[5]  
Gururangan S, 2020, Arxiv, DOI arXiv:2004.10964
[6]   Protected Health Information filter (Philter): accurately and securely de-identifying free-text clinical notes [J].
Norgeot, Beau ;
Muenzen, Kathleen ;
Peterson, Thomas A. ;
Fan, Xuancheng ;
Glicksberg, Benjamin S. ;
Schenk, Gundolf ;
Rutenberg, Eugenia ;
Oskotsky, Boris ;
Sirota, Marina ;
Yazdany, Jinoos ;
Schmajuk, Gabriela ;
Ludwig, Dana ;
Goldstein, Theodore ;
Butte, Atul J. .
NPJ DIGITAL MEDICINE, 2020, 3 (01)
[7]  
Sanh V, 2020, Arxiv, DOI [arXiv:1910.01108, DOI 10.48550/ARXIV.1910.01108]
[8]   Differences in the signs and symptoms of hyperthyroidism in older and younger patients [J].
Trivalle, C ;
Doucet, J ;
Chassagne, P ;
Landrin, I ;
Kadri, N ;
Menard, JF ;
Bercoff, E .
JOURNAL OF THE AMERICAN GERIATRICS SOCIETY, 1996, 44 (01) :50-53